Modeling And Analysis Of The Freeway Collection And Distribution Network Of Port Cluster With Shared Hinterlands

Global economy and trade is still in slow recovery from the 2007 global financial crisis. Given the big picture that developed countries were severely hit during the crisis, developing countries especially East-Asia emerging economies represented by China are now the growth engine for global economic recovery. During the period between 2007 and 2013, import and export volume of China recorded an average growth rate of 10.29%, 2.72% higher than world average level. Inside China, as one of the provinces where outbound economies and international logistics activities are most developed, Guangdong Province accounted for 1/4 of national import and export volume.As a key segment in the chain of international logistics, ports play as the hub of regional passenger, logistics, capital and information flow. Significant stimulation effects have been proved between ports and port cities in their regional economic development. Port cluster refers to an organic whole of several ports linked to each other by several modes of transportation in a specific region. Port cluster serves its economic hinterlands. When a hinterland is served by more than one port simultaneously, competition between ports for hinterland resources happens. The transportation network which carries logistics flows between hinterlands and port clusters are known as collection and distribution network. Among all transportation modes, highway transportation accounts for approximately 70% of overall transportation volume.Techniques and tools for the study of collection and distribution network are various. From the perspective of time-variant property, they can be divided into analytical methods and computer simulation methods. Analytical methods consist of complex network and goal programming etc. Structural and static analysis can be performed using these methodologies. On the other hand, computer simulation using simulation software such as Arena, can build up models capable of analyzing the dynamic properties of network. In the context of this paper, simulation methodology is useful in analyzing the dynamic characteristics of container cargo flows in the collection and distribution network. The simulation modeling in this paper is a multi-agent vehicle routing problem(VRP), which is an intelligent and dynamic decision in reality. To build the model of this paper, discrete-event simulation is more advantageous than agent-based simulation in the sense of model complexity. But the dynamic intelligent routing of cargo flow entities is difficult to be performed. Therefore, a dynamic intelligent routing algorithm is embedded in the discrete-event simulation model to meet the purpose. Comparisons between the outcomes of simulation and analytical methods can give us a more comprehensive understanding about the network.This paper builds a simulation model and performs analysis on the highway collection and distribution network in Guangdong Province. Insights and improvement advises are drawn from the perspective of logistics infrastructure policy and plan makers based on the analysis outcomes.